THE
MIT JAPAN
PROGRAM
(a4
- 9
,
..
Science, Technology,
Management
0
' 7Y_
iAE
TRANSFERRING JAPANESE
HUMAN RESOURCE PRACTICES:
JAPANESE AUTO PLANTS IN JAPAN AND THE U.S.
John Paul MacDuffie and Frits Pil
MITJP 96-13
Center for International Studies
Massachusetts Institute of Technology
Transferring Japanese Human Resources Practices:
Japanese Auto Plants in Japan and the U.S.
John Paul MacDuffie and Frits Pil
MITJP#96-1 3
Distributed Courtesy of the
MIT Japan Program
Science * Technology * Management
Center for International Studies
Massachusetts Institute of Technology
Room E38-7th Floor
Cambridge, MA 02139
phone: 617-253-2839
fax: 617-258-7432
O MIT Japan Program
______1_________1______________l___l
TRANSFERRING JAPANESE HUMAN RESOURCE PRACTICES:
JAPANESE AUTO PLANTS IN JAPAN AND THE U.S.
JOHN PAUL MACDUFFIE
FRITS PIL
IMVP RESEARCH BRIEFING MEETING
JUNE 1994
Transferring Japanese Human Resource Practices:
Japanese Auto Plants in Japan and the U.S.
John Paul MacDuffie and Frits Pil
Wharton School, University of Pennsylvania
January 1994
John Paul MacDuffie (presenter)
2017 Steinberg-Dietrich Hall
Wharton School, U. of Pennsylvania
Philadelphia, PA 19104-6370
(215) 898-2588 (phone)
(215) 898-0401 (fax)
Submitted to International Management Division
1_1__1
_
_Is
__s____l____l______l_1__111-
Frits Pil
2000 Steinberg-Dietrich Hall
Wharton School, U. of Penn.
Philadelphia, PA 19104-6370
(215) 898-1235 (phone and fax)
ABSTRACT
(50)
Transferring Japanese Human Resource Practices:
Japanese Auto Plants in Japan and the U.S.
We examine the transferability of human resource practices used by Japanese
manufacturers with 1993 data from Japanese-owned automotive assembly plants in Japan- and
the U.S. We find some well-known practices in Japan are not transferred to the U.S. while
most practices related to work organization are transferred, albeit with modifications.
___
I_
3
The structure and operations of Japanese manufacturing firms have evoked a flurry of
interest from researchers and business leaders the world over. As early as 1973, in-depth research
appeared contrasting the "Japanese production system" with the manufacturing practices of the West
(Dore 1973). The recent literature is replete with anecdotal and other evidence proclaiming the
wonders of Japanese manufacturing systems (see Young 1992). Furthermore, there is substantial
evidence that in several industries, these practices have been instrumental in allowing Japanese
plants to outperform their competitors in both productivity and quality (Abegglen & Stalk, 1985,
Womack et al., 1990, MacDuffie & Krafcik, 1992).
There are multiple elements that make the Japanese production system unique, including
product design practices, accounting control systems, and purchasing philosophies. We will focus
on another critical aspect of the Japanese production system -- namely, its human resource
practices, and the way these are used to develop and utilize a highly flexible, well skilled, and
motivated workforce (MacDuffie 1991). The idea that human resource policies can help firms
create competitive advantage is found not only in the literature about Japan, but also in an
extensive body of literature about the U.S. (e.g. Tichy et al. 1982; Dyer, 1988; Kochan, Katz, and
McKersie, 1986; Lawler, 1992). However, one crucial question raised when exploring the role of
human resource policies in the Japanese production system is whether or not these policies are
transferable outside the Japanese context.
If the world fit a true neoclassical economic model, there would be no difficulty in
implementing Japanese human resource systems abroad. It is clear, however, that there are
significant cultural and institutional differences that might slow or inhibit the adoption and
implementation of such practices (Schein, 1986). Westney (1987) argues that in the process of
____
LIII·-I--_ILII···i-5i·i
·-
__I___
_
I
4
cross-cultural transfer, practices get modified and adapted to fit the local environment. Kochan,
Dyer, and Batt (1992) state that understanding the transferability of practices across national
boundaries is one of the critical challenges facing researchers in the human resources/industrial
relations field.
The best way to study whether Japanese human resource practices are transferable, either
completely, or in modified form, is to look at Japanese plants that establish operations overseas,
and to analyze whether their human resource systems are the same in both locations. These firms
would be fully cognizant of how their human resource systems work in Japan, and therefore, lack
of transference could not be attributed to ignorance of how the system works.
It has long been acknowledged (e.g. Doeringer and Piore, 1971) that economic forces and
product market decisions affect the internal operations of the firm and how the employment
relationship is organized. We focus here on a single industry context -- automotive manufacturing - in order to avoid the confounding influence of different product markets and production
requirements that would be found in a cross-industry sample. The fact that Japanese automotive
companies have established a number of "transplant" manufacturing facilities in the U.S. makes this
a particularly relevant choice for studying the transferability of employment practices.
Much has been written about human resource practices in some of the Japanese automobile
transplants (e.g. Shimada and MacDuffie, 1986; Brown & Reich, 1989, Fucini & Fucini, 1990;
Adler, 1992), but no one has provided a detailed comparison of the full set of Japanese automobile
transplants in North America with a similar set of plants in Japan. Some authors (Kenney and
Florida, 1993; Young, 1992) have argued that Japanese auto companies have maintained their
manufacturing practices upon moving to the United States rather than modifying them to meet the
-
are
a
9F-
"-----------
I-
5
needs of the US environment. Others have argued that the Japanese production system, and its
human resource system in particular, is not well suited to the U.S. context and must been modified
extensively in order to function. (Parker and Slaughter, 1988) However, most of these arguments
are generalizations based on readings of work on particular transplants, or on what has appeared in
the popular press.
In order to understand whether Japanese human resource practices are transferable, and
indeed, whether they are critical for successful automobile manufacturing, it is important to look in
detail at the practices as they are put into effect in Japan. One body of literature has emphasized
the "so-called" three pillars of the Japanese employment relationship -- life-time employment,
seniority wages, and enterprise unionism (Shimada, 1985; Smith & Misumi 1989). As we will
show, evidence from the Japanese auto transplants indicates that these may not be as crucial to the
Japanese production system as they were once thought to be. Another body of literature explores
the human resource policies more directly related to work organization and skill development in the
Japanese production system, e.g. team based production methods, a small number of job
classifications, few distinctions between management and employees, and worker participation in
problem solving (Koike, 1989; Cole, 1979; MacDuffie & Krafcik, 1992). As we will show, these
latter practices are all prevalent in the transplants, but appear to undergo some adaptation to the
North American institutional and cultural environment.
Data and sample
All the major Japanese car manufacturers opened assembly plants in North America, either
wholly-owned or joint ventures with American or Japanese partners, between 1982 and 1992 --
1_1--_11_-·----
__
I
___1
____·_
.--
_I
_I I
_·11
6
Toyota, Nissan, Honda, Mazda, Mitsubishi, Fuji Heavy Industries (Subaru) with Isuzu, and Suzuki.
Our data are collected from an international survey of automotive assembly plants worldwide,
sponsored by MIT's International Motor Vehicle Program. Here we report plant-level survey data
on human resource practices from 8 of the 11 Japanese auto transplants in North America (survey
available from authors upon request). All survey responses are for the 1993 model year (i.e. from
Sept. 1992 to August 1993). In addition, we have performed extensive plant visits and interviewed
management representatives at five of these plants.
We have collected the same survey data for 12 automobile plants in Japan. We have done
extensive visits of 12 Japanese plants, and interviewed the management of 10 plants, as well as
corporate-level managers at three of the five major companies in Japan. The data we present below
are all based on the survey responses, but our commentary will reflect insights we have developed
during the course of visiting the plants and conducting our interviews. At this preliminary stage of
our data analysis, we present primarily comparisons of means between Japanese plants in the Japan
and those in the U.S.
Assessing the Transferability of Japanese Human Resource Practices
The first main-stay of the Japanese human resource system is its guarantee of permanent
employment for workers. This is important for the successful implementation of a whole range of
practices, including the provision of extensive training, successful team work, and employee
commitment to continuous improvement. This permanent employment is offered to a set of core
employees. Part-timers, seasonal employees, and contract workers are used to handle demand
fluctuations, and these do not receive employment guarantees (Dore, 1986). Today, these employees
-a
a
IWt----
7
make up about 6.5% of the workforce in the Japan-based plants that we surveyed. In the
transplants, they make up on average less than 1% of the workforce. Since this smaller "buffer" of
temporary employees gives the transplants much less protection from demand fluctuations (and
given the pronounced cyclicality of auto sales in the U.S.), one might expect it to be much more
difficult for the transplants to give employment security guarantees.
The two transplants that are unionized have included a commitment to employment security
in their union contracts. These do not constitute legal guarantees, because they are conditional on
such a policy not jeopardizing the financial viability of the company. The nonunion transplants
have no such formal agreements, but have also indicated a strong commitment to maintaining longterm employment for their core workers. None of the transplants have had any layoffs of these
core employees to date. They have been in a growth mode for much of their existence, and during
downturns, workers not needed for efficient production typically receive additional training.
However, the transplants make no employment commitment to their temporary workers. Mazda
and Diamond Star, for example, have already laid off some of these workers. (source?)
Closely related to the employment security issue is the question of whether employees
actually stay with the company once hired. Dore (1973), in his comparison of British to Japanese
factories, noted that in the Taga plant, the average age of the employees was 31.3 years. The
average age of entry was 21.8 years, and the average length of service was 9.5 years. This suggests
that on average, the Japanese employees had worked solely for their company since entering the
labor force. We find that this pattern persists to the present. The average age in the Japanese plants
is now 36.5 years, versus 34.5 in the transplants; while this difference is statistically significant, its
magnitude is not that great. However, the employees in Japan have been at their plant an average
_I111I·------I1--.iill
Xllniil(i-llI--.^a
I_
LLLII------- _
8
of 15.4 years versus the 4.2 years that the average transplant employee has worked at his/her plant.
However, it must be remembered that the transplants are extremely new, with half having started
operation in the last five years.
Turnover (defined as resignations, retirements, and terminations) is currently quite high in
Japan plants compared to the transplants, (8.7% vs 2.7%). The relatively high turnover in Japan is
surprising, given the long-term relationships the Japanese plants are purported to have with their
employees. Since the data are from 1992-93, a time of recession in the Japanese auto industry, it is
unlikely that the high turnover reflects the labor shortages common in manufacturing industries
during the boom years of the late 1980s. Instead, it may reflect the widely publicized reluctance of
young Japanese to stay in manufacturing jobs.
The low rate of turnover in the transplants is also a surprise, given the propensity for jobswitching in the U.S.
These trends suggest that the gap in employment tenure between the two
groups of plants will narrow further in years to come. Furthermore, the turnover rate combined
with the unscheduled absenteeism rate (which is also quite low at the transplants -- 2.2% compared
with 1.2% in Japan and 5.1 % for U.S.-owned plants) do not provide support the view that
American workers at the transplants are dissatisfied with their experience in a Japanese production
system, as predicted by Parker and Slaughter (1988).
Long term employment is very important when companies consider investment in training.
Firms that are planning to utilize employees for longer periods of time, and that are confident
employees won't leave, are more likely to invest in employee development than firms that treat
their workers as an expandable commodity. Given the greater emphasis at Japan plants on a longterm employment relationship, we would expect them to have higher training levels. The lower
.·*D-lnar--JII···---
I
turnover levels at the transplants, on the other hand, and their "start-up" status would lead us to
expect higher levels of training there.
In Table 1, we see that although both Japan plants and transplants provide similar levels of
training to new employees, the transplants provide significantly higher levels of training to
experienced employees. Among experienced employees, while there is not statistically significant
difference in the level of training for production workers, the transplants do provide much more
extensive training to experienced first-line supervisors and engineers. This may be due to the fact
that our "experienced" category encompasses all employees with more than one year of tenure. To
the extent that more training is needed during the initial 5-10 years on the job for these employees,
the relative newness of the transplants could explain this differential.
TABLE 1: TRAINING LEVELS
TRAINING TYPE:
JAPAN
New Employee Training
2.81
2.75
Experienced Employee Training
1.67
4.13
Training For New Hires By Category
New Production Worker
New Supervisor
New Engineer
2.91
2.91
2.82
3.00
2.88
2.88
Training for
Experienced
Experienced
Experienced
TRANSPLANTS'
Experienced Employees by Category
3.22
Production Worker
Supervisor
2.78
First-Line
2.00
Engineer
4.12
4.63
4.50
t-test
***
***
***
For new employees, 1= 0-40 hrs/ycar; 2= 41-80 hrs/year: 3=81-160 hrs/year 4= 160+ hrs/year.
For experienced employees. 1=1-20 hrs/year: 2=21-40 hrs/year; 3=41-60 hrs/year; 4= 61-80 hrs/year: 5=80+ hrs/year.
T-tests done using 2-sided confidence intervals: * = .1; ** = .05; *** = .01
We emphasized earlier the importance of having a good detailed understanding of the
..___.- ______
I__~
10
Japanese practices in Japan in order to study whether they are transferred abroad. This significance
is highlighted when we look in detail at the training for production workers. There is little
difference in the hours of training for production workers between the Japan plants and the
transplants. However, when we break down the training of new and experienced production
workers we see that there are significant differences in the type of training being provided to the
two groups of plants. We will refer to specific differences in production worker training below.
TABLE 2: BREAKDOWN OF TRAINING FOR PRODUCTION WORKERS
New Production Workers
Japan
Percent of Training Time Spent on:
Plant orientation
20.5
Basic skills
1.3
14.7
Interpersonal skills
5.2
SPC & problem-solving
2.8
Production methods & philosophies
30.0
General technical skills
3.8
Specific technical skills
13.7
Health & safety issues
1.1
Environmental policies
7.0
Other
Transplants
22.5
0.0
4.7
5.6
13.5
13.1
6.4
19.6
11.1
3.5
Experienced Production Workers
Japan
Percent of Training Time Spent on:
Plant orientation
3.9
1.7
Basic skills
24.8
Interpersonal skills
41.9
SPC & problem-solving
9.9
Production methods & philosophies
General technical skills
4.7
0.3
Specific technical skills
8.3
Health & safety issues
0.0
Environmental policies
Other
4.7
Transplants
0.0
0.0
4.6
12.4
26.6
10.1
8.8
23.0
6.7
7.9
T-tests done using 2-sided confidence intervals: *=.l, **=.05, ***=.01
-
.II^------
t-test
**
*
**
**
**
**
*
**
11
In addition to training, workers become multi-skilled through job rotation. Through
extensive rotation, workers familiarize themselves with a range of jobs. This helps eases scheduling
problems, helps with redeployment of workers as work loads shift, eases ergonomic problems from
repeated operations, helps reduce the monotony of work, and helps with employee problem solving
efforts (to be discussed below). It also helps prevent the development and assertion of sectional
interests that are inconsistent with the goals of the plant.
We have found that job rotation in the Japan plants is more extensive than in the transplants.
Although the transplant workers rotate some within their work group, they rarely rotate across work
groups. In the Japan plants, workers rotate within their work groups, and habitually rotate across
work groups as well. Florida and Kenney's work with transplant auto parts suppliers (1993)
suggests that this may be due to workers in U.S. plants not yet having the knowledge to move
regularly from team to team. Transplant workers rotate at more frequent intervals, however, than
workers in japan, typically for ergonomic reasons.
To help ease job rotation, pay in Japan plants is attached not to the job, but rather to
employee rank. Employees in a given rank can do a range of jobs, and rank is often more an
indicator of seniority and how many jobs an employee has learned than a "right" to work in a
particular job. These are analogous to job classifications in the U.S., although, as noted above, are
not necessarily attached to a particular job. Plant-specific customs dictate the general pattern of
rotation among jobs for employees of the same rank (Koike, 1989). The Japan plants in our sample
have five different ranks for production workers and five ranks for maintenance workers.
In comparison, the transplants in our sample have only one production worker classification,
and one or two maintenance worker classifications. This is very low for the U.S. context, where
-1 - 11----
12
traditional plants often had hundreds of classifications. As in Japan, pay is completely separated
from the actual job that is performed. In both sets of plants, this approach shifts employees'
incentives away from moving up a strict job ladder and towards skill and knowledge acquisition as
a source of status or satisfaction (MacDuffie, 1988).
Given the low number of ranks, how is employee compensation determined? According to
Aoki (1990), unions in Japan do not negotiate pay rates for each job. Rather, they negotiate about
entry-level wages, inter-rank wage differentials, and rates of inter-rank progression. However, given
the small number of ranks, this does not offer much leeway in what to pay employees. As Table 3
below shows, the Japan plants make extensive use of bonuses. These are provided based on plant
and company performance, but also based on individual performance.
Among the transplants, some give bonuses for plant or company performance, but none give
bonuses to production workers based on individual performance. ("Company" for many of the
transplants refers to the U.S. manufacturing subsidiary, and as such is synonymous with "plant",
except when the same parent company has two different transplant facilities.) Almost all the
transplants, however, provide merit increases in salary to engineers and supervisors. This is less
common in Japan because it is felt that it leads to a high fixed wage structure that cannot be
adjusted during downturns. It is also interesting to note that no transplant offers bonuses for
seniority. Since there is only one rank of production worker, unlike in the Japan plants, this means
that pay at the transplants bears no relationship to seniority. This shift away from seniority based
compensation may reflect the fact that Japanese companies are not happy with it. 72% of Japanese
managers report that the seniority system adversely affects the morale of those who are the most
able (Smith & Misumi, 1989).
_
a
------ -^---·---·---r-a.---·Ladrraral-----
13
Table 3: CONTINGENT COMPENSATION
Percent of Plants Reporting
Compensation Type
Bonus for Company Performance
Bonus for Plant Performance
Bonus for Individual Performance
Bonus for Seniority
Merit Increase in Salary for
Individual Performance
Prod. Workers
JPN
Trans
50%
63%
50%
38%
50%
0%
33%
0%
50%
0%
Supervisors
JPN
Trans
50%
63%
50%
25%
50%
13%
33%
0%
58%
88%
Engineers
JPN
Trans
42%
63%
17%
25%
17%
13%
25%
0%
25%
100%
A frequent claim regarding compensation in Japanese plants is that there is an emphasis on
minimizing the pay differentials between employees. This is said to enhance the sense of
community and equal status among employees at different levels (e.g. Womack et al., 1990, Florida
& Kenney, 1991). The literature on intra-company wage differentials has stressed the relatively
low differential between the highest and lowest paid employees in Japanese companies.
Our research indicates, however, that the differentials between the lowest and highest ranks
within employee categories is much higher in Japan than in the Japanese transplants in the US (see
Table 4). We have found that this differential is not due to differences in starting pay (2100 yen =
$1). There is no statistical difference in entry-level pay for production workers between Japanese
plants and the transplants, nor for engineers. The lowest paid supervisors in Japanese plants earn
more than their counterparts in the transplants. The reverse is true for maintenance workers.
Table 4: PAY DIFFERENTIALS
Pay Differential between
Highest and Lowest paid
JAPAN
Production Worker
Maintenance Worker
First-Line Supervisor
Manufacturing Engineer
204%
205%
117%
446%
TRANSPLANTS
24%
12%
31%
130%
t-test
***
***
***
***
T-tests done using 2-sided confidence intervals levels: *=. 1, **=.05, ***=.01
lls____
_UJ·C
14
Since pay in this case includes both base wages and various bonuses (while excluding
benefits), it seems likely that these very large differentials reflect the use of bonuses to reward
individual-level differences in seniority, skill, and initiative in Japan-located plants, compared to the
policy at the transplants of awarding bonuses equally to all employees in a given category.
The importance of team work in Japanese plants has long been recognized by scholars (Aoki
1990, Koike 1989). Like plants in Japan, the transplants make extensive use of teams.
Management generally appoints team leaders, although at the unionized transplants, union officials
are often involved in team leader selection. Teams in both Japan plants and transplants are similar
in the degree of influence they have in various areas, according to management respondents.
Teams are reported to have the most influence over who should do what job and methods of work,
and the least influence over the selection of team leaders and the amount and pace of work. The
only statistically significant differences between the two groups are in the area of employee voice,
with teams in the Japan plants reported to have more influence over performance evaluations and
settling grievances and complaints.
Table 5: TEAM INFLUENCE
AREA OF INFLUENCE
JAPAN
Use of new technology on the job
Who should do what job
The way work is done; revising methods
Performance evaluations
Settling grievances or complaints
How fast the work should be done
How much work should be done
Selection of team leader
TRANSPLANTS
2.3
3.6
3.6
2.7
3.5
2.3
2.0
1.3
2.0
3.1
4.1
1.4
2.1
2.0
1.6
2.1
Note: Scale from 1-5 where 1 = no influence, and 5 = extensive influence
T-tests done using 2-sided confidence interval: *=.l, **=.05, ***=.01
__ -------
-
~ 11-- --
-- `__
t-test
*
*
I
15
Although Japan plants and transplants both make extensive use of teams, they differ
significantly in the extent to which their employees engage in continuous improvement of the
production process (known as kaizen) through off-line problem-solving. One such activity is
quality circles. Looking at Table 6, we find that a much larger percent of the workforce in Japan
plants are found in quality circles or employee involvement groups than in the transplants (although
even among the transplants there is quite a bit of variance). One possibility is that kaizen leads
employees to fear that they will work themselves out of a job (Young, 1992). However, the
employment security assurances of the transplants are intended to address precisely those concerns.
An alternative view comes from Kenney and Florida (1993), who suggest that the low level of
quality circle and employee involvement activity in the transplants reflects their newness, and that
plants plan to increase their usage over time. Furthermore, according to Cole (1979), participation
in these "voluntary" small group activities in Japan plants is more likely to be viewed as mandatory
by employees, due to management and peer pressure, than in U.S. plants.
The Japan plants and the transplants differ significantly in how the quality circle activities
are organized. In Japan, a third of the plants reported having quality circles or employee
involvement groups meeting for at least a portion of their time during non-paid hours. All but one
of the transplants pays employees during the time their quality circle meets. The heavy emphasis
placed by the Japan plants on problem-solving activities is seen from table 2, which shows that
over 40% of the training provided to experienced production workers in those plants is related to
problem solving and SPC, compared to just over 10% in the transplants.
The Japan plants also receive significantly more suggestions per employee than do the
transplants. However, several of the plants we visited in Japan told us they place explicit quotas on
_
-.·T-e-
---·---*1II
___1_
__1I1I__I___I
-1411I1TYI
16
the number of suggestions employees should make each month. That is generally not the case in
the transplants. Only seven of the twelve Japanese plants responded to the question of whether they
paid for suggestions. However, all seven answered affirmatively. In contrast, only three of the
eight transplants paid for suggestions. We know from interviews that the transplants prefer nonmonetary forms of recognition, such as reward ceremonies or points that can be accumulated
towards prizes, because of concerns that monetary awards cause employees to focus solely on
major-impact suggestions that generate large awards. Furthermore, transplant managers believe that
it is difficult to establish an allocation system for monetary awards that will be perceived as
objective and equitable.
Table 6: KAIZEN ACTIVITIES
Indicators of Off-line
Problem Solving
JAPANESE
81
Percent of employees in Quality
Circle or Employee Involvement group
Suggestions per Employee
130
Percent implemented
80
TRANSPLANTS
t-test
27
***
4
60
**
T-tests done using 2-sided confidence intervals: *=.1, **=.05, ***=.01
Before the transplants opened, one common expectation about the transferability of Japanese
employment practices was that American workers would be too individualistic, too diverse, and too
poorly educated to fit with that system. Yet the transplants have been highly successful, in terms
of productivity and quality performance, with an American workforce whose turnover and
absenteeism is very low. One reason for this success is that the transplants spend inordinate
amounts of time on selecting and socializing their employees. The ten Japan plants in our sample
that hired production workers in the recent past hired on average over 80% of those who applied.
17
Only three of the transplants have hired production workers recently, but they were much more
selective.
On average, they hired only 5% of those who applied. Those who are hired are very
well educated (Table 7), with almost 40% of production workers having some college education.
Table 7: PRODUCTION WORKER EDUCATION LEVELS
Percent of Employees
by Educational Category
JAPANESE
TRANSPLANTS
No Secondary School
Some Secondary School
Secondary School Degree
Some College
College Degree
16.2
0.5
82.5
0.2
0.7
1.0
6.8
54.0
23.8
14.4
t-test
*
***
***
T-tests done using 2-sided confidence intervals: *=.1, **=.05, ***=.01
Some academics also argue that the transplants strove to approximate the homogenous labor
force they have access to in Japan by locating in states with few minorities (e.g. Cole & Deskins
1989). Although the locational choices of the Japanese plants may reflect some biases on the part of
the Japanese parent company, it is not clear to us that these biases persisted after the plants opened.
In every transplant we visited, we found minorities and women represented just as extensively as in
most American plants that we have visited. Indeed, we found that 19.5% of the production workers
in the transplants we surveyed are female, compared to only 11.6% in other North American plants,
and only 2.1% in Japanese plants. While the transplant workforces may be demographically
diverse, the extreme selectivity during the hiring process may mean that workers are homogeneous
with respect to attitudes towards work and receptiveness to Japanese manufacturing philosophies
and human resource practices. The high amount of training in Japanese production methods also
helps create a strong and consistent organizational culture (see Table 2).
One major institutional difference between the transplants and the Japan plants is the
,-1.
I-~,
*I I
"I.
I. TI
~
a
v-·--
-.--
I---
.
-
18
structure of worker representation. All the Japan plants in our sample are unionized with enterprise
unions.
Enterprise unions have long been viewed as crucial to the success of Japanese companies,
because the union represents both blue-collar and white-collar employees in a single firm, and thus
can help align employee goals with those of the organization. While Shimada (1985) disputes the
claim that enterprise unions are docile and always cooperative with management, it is clear that this
union structure is associated with a high level of management-union consultation on strategic issues
and a less adversarial relationship, given that many managers begin their careers as union members
in the lower management ranks.
In contrast, only two of the eight transplants in our sample are unionized, and their workers
are represented by the United Automobile Workers (UAW), an industrial union. There is some
evidence that the transplants have tried to create similar dynamics to those existing with an
enterprise union. At the unionized transplants, the labor contract includes a union commitment to
support the competitiveness of the plant (together with a management commitment about
employment security) and establishes a variety of mechanisms for ongoing labor-management
consultation. Five of the six nonunion transplants, in turn, have made efforts to establish some
governance structure for employee representation by establishing committees of worker
representatives (typically appointed by management) to provide worker representation. However,
transplant managers discuss fewer issues with these employee committees than do managers at the
Japan plants. As is shown in Table 8, the only issue most of the transplants discuss with their
employees are health and safety related. The other issues are generally only discussed by the two
unionized transplants, with only one non-unionized transplant discussing grievance issues, personnel
policies and employment conditions, and changes in strategy or policy, with their worker
_I
19
representatives.
Table 8: TOPICS THAT ARE DISCUSSED WITH WORKER REPRESENTATIVES
Issues Addressed by
Worker Representatives
(Percent of firms responding)
Grievances/Due Process
Personnel Policies/Employment Conditions
Health & Safety
Major Changes in Strategy or Policy
JAPANESE
100.0
66.7
100.0
58.3
TRANSPLANTS
37.5
37.5
87.5
25.0
When asked about the influence of worker representatives on management decisions (Table
9), transplant managers claimed the influence was greatest for decisions about new technology and
work allocation. In contrast, managers in the Japan plants indicate that worker representatives have
a lot of influence over decisions about reductions in staffing and changes in the amount of overtime
to be worked, but relatively little in other areas.
Table 9: WORKER INFLUENCE ON DECISION MAKING
Influence Exerted by Worker
Representatives before Formal
J APANESE
Management Decision
1.8
(e.g.
plans
investment
New
expansion of current operations)
1.0
Design of new technology
1.0
Evaluation and selection among
new technology options
1.8
Restructuring of jobs and duties
as result of new technology
1.4
Planning and coordination of
training for new technology
4.9
Reductions in staffing
1.9
Changes in work allocation
3.9
Amount of overtime to be worked
1.4
Changes in work to be subcontracted
TRANSPLANTS
1.6
t-test
1.6
2.2
2.6
2.6
2.2
3.0
1.8
2.0
*
or**li
T-tests done using 2-sided confidence interval: *=.1, **=.05, ***=.01
--
_________I
11
20
This analysis suggests that not all the characteristics of the Japanese human resource system
are necessary to achieve superior performance. Indeed, what were once considered the "three
pillars" of the Japanese human resource system: guaranteed lifetime employment, seniority wages,
and enterprise unionism, are not found in Japanese transplants. Why then, do the transplants
perform almost as well as their counterparts in Japan (MacDuffie & Krafcik, 1992; Womack et al,
1990)? We would argue that the key is a system of human resource practices -- those reviewed
above -- that result in a highly committed, motivated, and flexible workforce. This is not to say
that these practices have been transferred blindly to the transplants. Almost each one of them has
been modified or adapted in some way by the transplants, but generally in ways that are still
consistent with the overall logic of the production system.
It is harder to predict whether Japanese human resource systems can be adapted and
incorporated into the operations of existing U.S. plants, the "brownfield" sites. All of the Japanese
transplants, with the exception of NUMMI, were established at greenfield sites. Even at NUMMI,
the entire human resource system was rebuilt from scratch because of the plant closure and the new
management team. According to Koike (1989), production systems represent complex interactions
of technological and social systems (including human resource systems), and although it is easy to
change the technical systems, it is much harder to alter the social systems that develop in work
sites over time. We have shown that the transplants, which did not have to deal with entrenched
employment practices, have been successful in transferring and adapting Japanese human resource
practices to the North American environment. We are now in the process of conducting research to
see whether these practices can be transferred with the same level of success to older, more
traditional, unionized American plants.
_
llm_______·__1___1___11__
__
___
21
REFERENCES
Adler, P. 1992. "The Learning Bureaucracy: New United Motor Manufacturing Inc." in eds. B.
Staw and L. Cummings, Research in Organizational Behavior, Greenwich, CT: JAI Press.
Aoki, M. 1990. "Toward an Economic Model of the Japanese Firm." Journal of Economic
Literature. 28(1): 1-27.
Abegglen, J. & G. Stalk. 1985. Kaisha -- The Japanese Corporation. New York: Basic Books.
Brown, C. and Reich, M. 1989. "When Does Union-Management Cooperation Work? A Look at
NUMMI and Van Nuys." California Management Review 31(4): 26-44.
Cole, R. 1979. Work, Mobility, and Participation: A Comparative Study of Japanese and American
Industry. Berkeley: University of California Press.
Cole, R. and Deskins, D. 1989. "Racial Factors in Site Location and Employment Patterns of
Japanese Auto Firms in America. California Management Review 32(1): 9-22.
Doeringer, P., and M. Piore. 1971. Internal Labor Markets and Manpower Analysis. Lexington,
MA: Heath.
Dore, R. 1973. British Factory, Japanese Factory; The Origins of National Diversity in Industrial
Relations. Berkeley, University of California Press.
Dore, R. 1986. Flexible Rigidities: Industrial Policy and Structural Adjustment in the Japanese
Economy, 1970-1980. Stanford, CA: Stanford University Press.
Dyer, L., ed. 1988. Human Resource Management: Evolving Roles and Responsibilities.
Washington D.C.: BNA Books.
Florida, R. and M. Kenney. 1991. "Transplanted organizations: The transfer of Japanese industrial
organization to the U.S." American Sociological Review 56: 381-398.
Fucini, J., and S. Fucini.1990. Working for the Japanese -- Inside Mazda's American Auto Plant.
New York: Free Press.
Kenney, M. and Florida, R. 1993. Beyond Mass Production: The Japanese System and Its Transfer
to the U.S.. New York: Oxford University Press.
Kochan, T., L. Dyer, and R. Batt. 1992. "International Human Resource Studies: A Framework for
Future Research." in eds. David Lewin, Olivia Mitchell, and Peter Sherer, Research Frontiers in
Industrial Relations and Human Resources, Industrial Relations Research Association Series,
Madison WI.
_--i.l----
22
Kochan, T., H. Katz, and R. McKersie. 1986. The Transformation of American Industrial Relations.
New York: Basic Books.
Koike, K. 1989. Understanding Industrial Relations in Modem Japan. Translated by Mary Saso.
New York: St. Martin's Press.
Lawler, E.E. III. 1992. The Ultimate Challenge. San Francisco, CA: Jossey-Bass.
MacDuffie, J.P. 1988. "The Japanese Auto Transplants: Challenges to Conventional Wisdom."
ILR Report., 26(1): 12-18.
MacDuffie, J.P. 1991. Beyond Mass Production: Flexible Production Systems and Manufacturing
Performance in the World Auto Industry Ph.D. dissertation, Sloan School of Management, M.I.T.,
Cambridge, MA.
MacDuffie, J.P., and J. Krafcik. 1992. "Integrating Technology and Human Resources for High
Performance Manufacturing: Evidence from the International Auto Industry." In Transforming
Organizations, eds. Thomas A. Kochan and Michael Useem. New York: Oxford University Press.
Parker, M. and J. Slaughter. Fall 1988. "Managing by Stress: The Dark Side of Team Concept."
ILR Report., 26(1): 19-23.
Schein, E., 1985. Organizational Culture and Leadership. San Francisco, CA: Jossey Bass.
Shimada, H. 1985. "The Perceptions and Reality of Japanese Industrial Relations." in ed. Lester
Thurow, The Management Challenge: Japanese Views. Cambridge, MA: MIT Press.
Shimada, H. and J. P. MacDuffie. 1986. "Industrial Relations and Humanware: Japanese
investments in Automobile Manufacturing in the United States." Working Paper, MIT Sloan School
of Management.
Tichy, N., C. Fombrun, and M. Davanna. 1982. "Strategic Human Resource Management." Sloan
Management Review 2: 47-61.
Westney, E. 1987. Imitation and Innovation: Transfer of Western Organizational Patterns to Meiii
Japan. Cambridge, MA: Harvard University Press.
Womack, J., D. Jones, and D. Roos (1990). The Machine That Changed the World. New York, NY:
Rawson Associates.
Young, M. 1992. "A Framework for Successful Adoption and Performance of Japanese
Manufacturing Practices in the United States." Academy of Management Review, 1992, 17(4):
p.677-700.
.-------------